This article is mainly to help readers who are not very familiar with the physical principles of industrial CT understand the impact of industrial CT technical parameters on performance indicators, so that when selecting and purchasing industrial CT equipment, they can appropriately put forward technical requirements and reasonably to achieve a compromise between performance and price.
1 Basic characteristics of industrial CT
1.1 Overview of industrial CT
CT is computed tomography technology, which is the abbreviation of English Computed Tomography. The word tomography comes from the Greek word tomos, which means an X-ray photography technology that can take pictures of a single plane while removing the influence of other plane structures. Using the traditional human body perspective method, the three-dimensional human body is compressed into a two-dimensional image along the direction of X-rays. All bone structures and tissues in the body are overlapped, which greatly reduces the clarity of the object of interest. Thus, although it has excellent spatial resolution (the ability to distinguish closely adjacent high-contrast objects), it ends up with poor low-contrast resolution (the ability to distinguish low-contrast objects from the background). This has led to the emergence of traditional tomography technology [8].
The basic principle of traditional tomography is shown in Figure 1. Consider first two isolated points A and B in the patient's body: point A is on the focal plane and point B is outside the focal plane. The shadows projected by points A and B onto the X film are marked A1 and B1 correspondingly, as shown in Figure 1(a). At this time, the image generated on the film is completely indistinguishable from traditional photography, and then the X-ray source and the X-film are moved in opposite directions synchronously (for example, as shown in the figure, the X-ray source moves to the left and the X-ray film moves to the right) to the Two positions. We want to ensure that the shadow A2 generated by fixed point A coincides with the shadow A1 generated by point A at the first position. This is easily accomplished by setting the distances between the X-ray source and the X-film movement so that they are proportional to the corresponding distances to point A, as shown in Figure 1(b). However, the shadows B2 and B1 generated by fixed point B at the second position do not coincide with each other. This is because point B is not on the focal plane, and the distance ratio from point B to the X-ray source and point B to the film deviates from the corresponding distance ratio to point A. When the X-ray source and film move continuously along a straight line (naturally in opposite directions), the shadow generated by point B forms a straight line segment. This property is applicable to any point above and below the focal plane. It should be noted that points that are out of focus produce reduced shadow intensity because the shadow is spread over an expanded area. All points on the focal plane maintain the original image position on the film, their shadow is still a point, and the corresponding intensity does not decrease.
Figure 1 The principle of traditional tomography
Although this tomography technique has had some success in producing clear images of the plane of interest, they do not increase the contrast of the object. Nor can other structures outside the focal plane be fundamentally removed. The quality of the image is obviously compromised.
Modern tomographic imaging technology, namely CT, is a method based on the application of computers to reconstruct images from multiple projection data. In the modern tomographic imaging process, only thin layers passing through a specific section (the object being detected, The projection data of the section (or called slice) is used to reconstruct the image of the section, thus fundamentally eliminating the interference of other structures outside the "focal plane" of traditional tomography on the section of interest. The contrast has been significantly enhanced; at the same time, the image intensity (grayscale) value in the tomographic image can truly correspond to the radiation density of the material being inspected, and small changes in the radiation density within the object being inspected can be found. In fact, low-contrast detectability (LCD) is the key difference between CT and conventional radiography. This is also the most important factor in the rapid clinical acceptance of CT.
It needs to be emphasized that all non-destructive testing technologies except CT technology do not have this capability. Because there are no interference from overlapping structures, image interpretation is much easier than with traditional radiography.
New buyers can quickly understand CT results. Therefore, CT technology has been developing rapidly since the world's first medical CT scanning equipment appeared at EMI in the UK in the early 1970s. CT has now become one of the most commonly used clinical diagnostic tools. In recent years, the emergence of spiral CT has made this technology a big step forward.
The basic principle of industrial CT is the same as that of medical CT, so it also has all the basic characteristics of medical CT. The detection image has no interference from structural materials other than the detected "slice" and can detect extremely small material density changes within the detection object. At the same time the interpretation of the images is much easier than with conventional radiography.
Therefore, industrial CT is also widely used to check the internal structure or assembly correctness of mechanical parts, and can also be used to non-destructively measure the internal dimensions of parts. In recent years, in view of the fact that a large number of studies on various other non-destructive testing methods have not produced satisfactory results, industrial CT is considered to be the most promising method for inspecting drugs and explosives.
It is worth noting that CT detection obtains a radiation density distribution image, which more professionally should be called a distribution image of ray linear attenuation coefficient. Since in most cases there is an approximate correspondence between radiation density and material density, people often mistake CT images for general (material) density distribution images. This confusion is not very harmful in many practical applications, but it may lead to some illusions when accurately quantifying the test results.
Due to the different detection objects, industrial CT and medical CT are so different that there are almost no similarities in appearance. The detection object of medical CT is basically the human body or organs, and the range of changes in material density and external dimensions is relatively small. However, the detection objects of industrial CT are much wider, ranging from micron-level integrated circuits to large workpieces exceeding one meter, from wood or other porous materials with a density lower than water to heavy metal materials with high atomic numbers. They are all CT detection objects; The inspection requirements of interest vary from various types of internal defects to assembly structures and dimensional measurements. This makes the radiation sources, radiation detectors and system structures used in industrial CT systems for different purposes very different, and even the shapes of industrial CT systems are also very different. In this sense, understanding industrial CT may be more difficult than understanding medical CT.
The disadvantage of industrial CT is that its technology is complex and the equipment price is relatively high. The use and maintenance of the equipment are also relatively difficult. In addition, the amount of data required to reconstruct tomographic images is huge and the detection speed is slow.
1.2 The main components of industrial CT and their characteristics
An industrial CT system should at least include a ray source, a radiation detector, a sample scanning system, a computer system (hardware and software), etc. .
1.2.1 Types of ray sources
X-ray machines and linear accelerators are commonly used as ray sources, collectively referred to as electron radiation generators. In principle, the electron cyclotron can be used as a radiation source for CT, but due to its low intensity, it has rarely been put into practical use. The peak ray energy and intensity of the Although higher energies are achievable, they are mainly used only for experiments. The unique advantage of the electronic radiation generator is that it no longer produces rays after cutting off the power supply. This inherent safety is very beneficial for industrial site use. The focal spot size of an electron radiation generator is from a few microns to a few millimeters. In the process of converting high-energy electron beams into X-rays, only a small part of the energy is converted into X-rays, and most of the energy is converted into heat. The smaller the focus size, the greater the local power density on the anode target, and the higher the local temperature. . The actual applied power is determined based on the power density that the anode target can tolerate for long-term operation. Therefore, the power or maximum voltage of a radiation source with a small focus or even a micro focus is lower than that of a radiation source with a large focus.
The biggest disadvantage of electronic radiation generators is the pleochroism of the X-ray energy spectrum. This continuous energy spectrum of X-rays will cause spectrum hardening during the attenuation process, leading to various hardening-related artifacts.
The biggest advantage of the isotope radiation source is that its energy spectrum is simple, it consumes very little power, the equipment is small and relatively simple, and the output is stable. However, its disadvantage is that the intensity of the radiation source is low. In order to increase the intensity of the source, the volume of the source must be increased, resulting in an increase in the size of the "focus". There are few practical applications in industrial CT.
Synchrotron radiation is originally a continuous energy spectrum. Through monochromator selection, directional, almost mono-energy, high-intensity X-rays can be obtained, so a CT system with high spatial resolution can be made. However, since the ray energy is 20KeV to 30KeV, it can actually only be used to detect small samples of about 1mm and is used in some special occasions.
1.2.2 Radiation detectors
There are two main types of detectors used in industrial CT - discrete detectors and area detectors
1.2. 2.1 Discrete detector
Commonly used X-ray detectors include gas and scintillation detectors.
The gas detector has natural collimation characteristics, which limits the influence of scattered rays; there is almost no interference; and the device has good consistency. The disadvantage is that the detection efficiency is not easy to improve, and high-energy applications have certain limitations; secondly, the distance between detection units is several millimeters, which is too large for some applications.
Scintillation detectors are more widely used. The photoelectric conversion part of the scintillation detector can choose a photomultiplier tube or a photodiode. The former has an excellent signal-to-noise ratio, but due to the large size of the device, it is difficult to achieve a high level of integration and the cost is high. The most widely used in industrial CT is the scintillator-photodiode combination.
The main advantage of a discrete detector using scintillator is that the depth of the scintillator in the ray direction can be unrestricted, so that most of the incident X photons are captured and the detection efficiency is improved. Especially under high-energy conditions, the acquisition time can be shortened; because the scintillator is independent, there is almost no optical interference; at the same time, there are tungsten or other heavy metal spacers between the scintillator, which reduces X-ray interference. If the spacer is extended forward to form a collimator, it can also block scattered X-rays; the discrete detector can achieve a dynamic range of 16 to 20 bits without performance degradation due to scattering and interference. The readout speed of discrete detectors is very fast, on the order of microseconds. At the same time, the accelerator output pulse can be used to gate the data collection to minimize the noise superimposed on the signal. Discrete detectors are also the least sensitive to radiation damage.
The main disadvantage of discrete detectors is that the pixel size cannot be made too small, and their adjacent intervals (pitch) are generally greater than 0.1mm; in addition, the price is more expensive.
There are some reports on the use of CdZnTe semiconductor detector arrays in industrial CT. Semiconductor detectors are commonly known as solid ionization chambers. Since they are sensitive to X-rays, there is no need for additional scintillator. The size of this detector can be made smaller and there is no optical interference. If there are no heavy metal spacers between detection units, the influence of scattered X-rays cannot be avoided. It should be said that this is a CT detector with great application prospects, but there are still some technical problems that need to be solved, such as excessive afterglow.
1.2.2.2 Area detectors
There are three main types of area detectors: high-resolution semiconductor chips, flat-panel detectors and image intensifiers. Semiconductor chips are divided into CCD and CMOS. CCD is not sensitive to X-rays, and the surface is covered with a layer of scintillator to convert X-rays into CCD-sensitive visible light. Flat panel detectors and image intensifiers also inherently require internal scintillator to first convert X-rays into visible light in the sensitive wavelength bands of these devices.
The semiconductor chip has the smallest pixel size and the largest number of detection units. The pixel size can be as small as about 10 microns. The number of detection units depends on the maximum size of the silicon single crystal, which is generally more than 50mm in diameter.
Because the detection unit is small and the signal amplitude is also small, several detection units can be combined in order to increase the measurement signal. In order to expand the effective detector area, lenses or optical fibers can be used to optically couple them to a large-area scintillator. The effective area of ??the detector can theoretically be extended to any desired length in one direction using fiber coupling. The use of optical coupling technology can also keep these semiconductor devices away from direct irradiation of X-ray beams to avoid radiation damage.
Semiconductor chips can also be used to form a line detector array. There is no isolation between the scintillator corresponding to each detection unit or a whole line of scintillator is covered on many detection units, which has the basic characteristics of an area detector. , apart from the advantage of small pixel size, its performance cannot be compared with discrete detectors. The image intensifier is a traditional area detector and a vacuum device. The nominal pixel size is <100μm, and the diameter is 152~457mm (6~18in). The readout speed can reach 15-30 frames/s, making it the fastest surface detector. Due to the inherent noise generated by statistical fluctuations during the image enhancement process, the image quality is relatively poor. Generally, the radiographic sensitivity is only 7 to 8. When a computer is used for data superposition, the radiographic sensitivity can be increased to more than 2. Other disadvantages are fragility and image distortion.